SIMULATION FOR ALD March 25, 2021 | online
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
© Dr. Xiao Hu, TU Chemnitz Live Web-Event SIMULATION FOR ALD March 25, 2021 | online COMMUNICATION is important! This Live Web-Event shall give you the opportunity to get into discussion and get to know each other. Listen to speakers and ask questions. But also discover the possibilities to speak with other participants in open or direct discussion. 1
Introduction Atomic layer deposition (ALD) has advanced tremendously in the past two decades into a > USD 3 billion annual market in equipment and chemicals. New integration schemes and patterning technologies are driving the need for atomistic modeling for materials and processes used for manufacturing computer logic and memory chips. The use of atomistic modeling in the semiconductor industry is at its infancy and some equipment and chemical suppliers have advanced more than others using the most recent atomistic modeling software tools available on the market. Other industries with high ALD engagement and interest in modeling include MEMS & sensors, battery technology, medical, display, lightning, barriers and photovoltaics. For a long time equipment companies and fabs have been using process chamber fluid flow simulations to optimize processes and ALD reactors for high productivity and product yield. The recent trends in 3D integration with smaller critical dimension and high aspect ratio features demand also so-called feature scale simulations. Very few research groups and leading industry players have been able to join all three scales of the ALD process into one common tool box, i.e., atomic scale, feature scale and reactor scale modeling. In this workshop the current state of research for modeling approaches on different length scales will be presented. Together we will discuss how to move forward to a multi-scale approach for ALD and related methods like atomic layer etching (ALE) and chemical vapor deposition (CVD). This workshop provides the opportunity to get in contact with industrial and academic partners, to learn more about fundamentals of ALD modeling and to get informed about recent progress in the field. If you are interested to join the online market place, please contact us directly. Process optimization using fluid dynamics simulation Poor process configuration improved process configuration © Linda Jäckel, Fraunhofer ENAS 2
PROGRAM COMMITTEE PROGRAM Linda Jäckel, Fraunhofer ENAS, Chemnitz, 12:30 WELCOME & MARKET PLACE Germany Discover the event website and meet other participants. Dr. Jonas Sundqvist, BALD Engineering 13:00 WORKSHOP SESSION 1 15:30 WORKSHOP SESSION 2 AB, Värmdö, Sweden Dr. Katrin Ferse, EFDS e.V., Dresden 13:00 Introduction 15:30 Presentation "Overview of simulation concepts "The DSMC Method - A for atomic layer processing" Simulation Approach for Low Jörg Schuster, Fraunhofer ENAS, Pressure ALD Processes" Chemnitz, Germany Asim Mirza, boltzplatz GmbH, Stuttgart, Germany 13:20 Keynote Presentation "Concepts and Preconceptions - 15:50 Presentation the Outlook for Simulations in "Modeling and Simulation of ALD © Dr. Xiao Hu, TU Chemnitz ALD" in a Level Set Framework" Simon Elliott, Schrödinger Inc., Lado Filipovic, Institute for New York, USA Microelectronics, Technische Universität Wien, Vienna, Austria 13:50 Presentation "Atomistic simulations of atomic 16:10 Presentation layer processes" "3D Monte Carlo simulations for Xiao Hu, Center for a better understanding of Microtechnologies, Technische conformality of ALD" Universität Chemnitz, Chemnitz, Véronique Cremers, Department ORGANIZER Germany of Solid State Sciences, Ghent European Society of Thin Films University, Ghent, Belgium 14:10 Presentation Gostritzer Str. 63 "In silico ALD – more bits and less 16:30 Presentation 01217 Dresden, Germany pieces" "Computational modelling of ALD www.efds.org Henrik Pedersen, Linköping reactors: Linking transport University, Linköping, Sweden phenomena with surface mechanisms" 14:30 BREAK Giorgos Gakis, National Technical Co-ORGANIZERS 14:45 Simulation Talk University of Athens, Athens, Fraunhofer-Institut für Elektronische moderated by Jonas Sundqvist Greece Nanosysteme ENAS and Linda Jäckel 16:50 Presentation Technologie-Campus 3 Experts - Simon Elliott, "Simulation for design of ALD 09126 Chemnitz, Germany Schrödinger Inc. and Jacques equipment" Kools, Encapsulix www.enas.fraunhofer.de/en Jacques Kools, Encapsulix S.A., 15:15 BREAK Simiane-Collongue, France BALD Engineering AB 17:15 END OF OFFICIAL PROGRAM Kanotvägen 4 17:15 - 17:45 13955 Värmdö, Sweden After Work Discussion www.blog.baldengineering.com moderated3 by Jonas and Linda 3
PRESENTATION Dr. Jörg Schuster Fraunhofer Institute for Electronic Nano Systems (ENAS) | Group leader simulation Dr. Jörg Schuster studied physics (diploma 1997, PhD 2002) at Chemnitz University of Technology. Since 2009 he leads the modeling and simulation group at Fraunhofer ENAS. His main working field is the simulation of devices, processes, and materials for micro and nano electronics. One of his special research interests is the application of atomistic simulation methods. He is author or coauthor of more than 90 publications in international journals and conference proceedings. "Overview of simulation concepts for atomic layer processing" March 25, 2021 | 13:00 – 13:20 This presentation will give a quick overview on simulation concepts and methods, which are in use along with atomic layer processing. There is a broad variety of concepts and methods available covering aspects of atomic layer processing from the molecular up to the equipment scale. This overview will explain its possibilities, assets, and limitations. Following the focus of the workshop, I will highlight the benefits of simulation methods from an application perspective. Simulation of ALD is useful for equipment manufacturers, people working with atomic layer process integration, or chemical companies developing new or improved precursors. By using computer simulations, we can analyze, visualize and optimize gas flows inside the chamber, identify critical process regimes or optimize process speed. I will illustrate this by a few application highlights from literature and our own research at Fraunhofer ENAS. 4
KEYNOTE PRESENTATION Dr. Simon Elliott Schrödinger Inc. | Director – Atomic level process simulation Dr Simon Elliott is a director at scientific software company Schrödinger. From 2001- 2018 he led a research group at Tyndall National Institute, Ireland. Prior to that he studied chemistry in Trinity College Dublin and Karlsruhe Institute of Technology. He qualified as a Project Management Professional and is a Fellow of the Royal Society of Chemistry. He was co-chair of the 16th International Conference on Atomic Layer Deposition and chair of a 175-member COST network on the same topic. "Concepts and Preconceptions - the Outlook for Simulations in ALD " March 25, 2021 | 13:20 – 13:50 Simon D. Elliott, David J. Giesen, Yuling An, Anand Chandrasekaran, Asela Chandrasinghe, Matthew Lawson, Mathew D. Halls Atomistic simulations in the field of atomic layer processing are introduced by looking at the key concepts - accuracy, speed, complexity, expertise, value - and our perceptions about them. We give recent examples of molecular and surface simulations that may challenge or reinforce those preconceptions. The success of deposition and etch processes depends crucially on the properties of the precursor chemicals used, so that much innovation in this field centers around developing new precursors. The space of possible ligands is vast, much too large to be explored systematically in the lab. The area is therefore ripe for computational screening and design, as long as the relevant properties can be computed quickly and accurately. We discuss the quantifiable metrics for screening. For instance, a good precursor should show wide temperature windows for delivery and deposition, which can be related to reactivity and stability, which can in turn be computed to some extent at the molecular scale. We illustrate the points with a computational screening of zirconium precursors against thermal stability and nitride formation. We also consider precursor volatility as a screening criterion and what makes it so difficult to simulate. Recent progress is reported in identifying its molecular basis and developing volatility models with machine learning, though many challenges remain. More generally, we discuss the emerging field of data science and potential benefits to the community if data can be better shared, curated and accessed. 5
PRESENTATION Dr. Xiao Hu Center for Microtechnologies, Technische Universität Chemnitz | Scientific employee Dr. Xiao Hu studied microelectronics at Technische Universität Chemnitz, where he obtained a Ph.D. degree in 2017 under the supervision of Prof. Schulz and Prof. Geßner. His dissertation was devoted to the multiscale simulation of copper ALD. Since 2017 he has been a scientific employee at Center for Microtechnologies, Technische Universität Chemnitz. Currently, he serves as the principal investigator of a DFG project with focus on the rational design of selective thermal atomic layer etching processes. His research concentrates on the surface chemistry of thin film deposition and etching processes. He is also working on the development and application of atomic and molecular simulation methods. "Atomistic simulations of atomic layer processes" March 25, 2021 | 13:50 – 14:10 With the rapid improvement of computer performance, atomistic simulations have become a powerful tool to understand and predict material properties. In the lecture, we will illustrate the use of atomistic simulations to investigate the chemical mechanisms of atomic layer processes, in particular ALD and thermal ALE. Different simulation methods, including density functional theory, molecular dynamics, and kinetic Monte Carlo simulations, will be introduced and discussed. The combination of these methods provides detailed information for atomic layer processes over a wide range of length and time scales. Based on examples of metal and metal oxide ALD, we will demonstrate how the ALD reactions can be predicted using atomistic simulations. Most of the predicted results are validated and directly compared with experimental data. Thermal ALE has emerged as a new technique for the isotropic etching of materials with atomic-scale precision. Here, we chose thermal Al2O3 ALE using HF and Al(CH3)3 as a case study. Based on a mechanistic understanding of ALE reactions, we will discuss the strategies for the design and screening of precursors. 6
PRESENTATION Prof. Henrik Pedersen Linköping University | Professor Henrik Pedersen received his M.Sc. in Chemistry in 2004 and his Ph.D. in Materials Science in 2008, both from Linköping University in Sweden. After a stint as industrial researcher at Sandvik Tooling Research and Development center in Stockholm, Sweden, he returned to academia and is today Professor of Inorganic Chemistry at Linköping University. His research is focused on understanding and developing new and better chemical vapor deposition methods by considering the time dimension in CVD, developing new precursors and new ways to use plasmas in CVD. “In silico ALD – more bits and less pieces” March 25, 2021 | 14:10 – 14:30 ALD presents a challenging, multi-scale modelling problem spanning at least ten orders of magnitude in both time and dimension; from the Ångström and picosecond scale of the surface reactions to the meter and seconds scale of the gas flows in the reactor. To establish a predictive modelling method for ALD, quantum chemical and DFT modelling of the surface interactions is connected to the gas flow patterns and gas exchange in the reactor via Computational Fluid Dynamics (CFD). We have recently shown how this connection is made in a continuous CVD process.[1] As the gas phase chemistry is (in the ideal case) negligible in ALD, it is a somewhat different process to model compared to CVD. Multiscale modelling of ALD is hampered due to the lack of CFD studies of ALD. We will discuss how predictive in silico ALD modelling could be done at the wafer scale. [1] Danielsson et al. A Systematic Method for Predictive In Silico Chemical Vapor Deposition. J. Phys. Chem. C 2020, 124, 7725. 7
PRESENTATION Asim Mirza boltzplatz GmbH | Co-Founder Dr. Asim Mirza studied aerospace engineering at the University of Stuttgart. He completed his PhD at the Institute for Space Systems, during which he specialized in the simulation of gas flows on a microscopic level and became one of the first developers of PICLas, an open source particle simulation software. Now, he is co-founder of the company boltzplatz GmbH which supports companies that tackle complex physical conditions with numerical simulations of rarefied gas and plasma flows. " The DSMC Method - A Simulation Approach for Low Pressure ALD Processes " March 25, 2021 | 15:30 – 15:50 The Direct Simulation Monte Carlo (DSMC) method is a powerful simulation model that treats gas flows on a microscopic level. It models gas flows using simulation particles that represent a large number of real molecules in a probabilistic approach to solve the Boltzmann equation approximately. Hereby, no continuum assumptions are made, so that rarefied gas flows can be simulated correctly even under non-equilibrium conditions. The DSMC method has been used in a wide range of engineering challenges ranging from atmospheric re-entry missions to flows in micro-electro- mechanical systems. An introduction to this method is given in this talk. It is outlined how the DSMC method could be utilized to simulate flows in low-pressure ALD processes, where classical CFD approaches are invalid. Additionally, the new open source software PICLas, which includes a state-of- the-art DSMC module, is introduced briefly. 8
PRESENTATION Prof. Lado Filipovic Institute for Microelectronics, Technische Universität Wien | Assistant Professor Lado Filipovic is an Assistant Professor of Modeling and Simulation of Integrated Semiconductor Sensors at TU Wien, where he obtained his doctoral degree (Dr.techn.) and venia docendi (habilitation) in 2012 and 2020, respectively. He is actively involved in several IEEE sponsored conferences, such as IEEE Sensors and SISPAD. His primary research interest is studying the operation, stability, and reliability of novel semiconductor-based sensors using advanced process and device TCAD approaches. "Modeling and Simulation of ALD in a Level Set Framework" March 25, 2021 | 15:50 – 16:10 The level set method, combined with Monte Carlo ray tracing, is a frequently used technique, when modeling surface transport in process technology computer aided design (TCAD). Using this method surfaces and interfaces, which move during deposition and etching, are defined implicitly by storing the distance to the desired surface on a regular grid. Modeling a multi-step process such as ALD using this technique requires the preservation of surface information during the course of the simulation of topography motion, which is not straight-forward since the explicit access to surface elements and nodes is not available. In this talk, the implementation of this framework in our in-house tool ViennaTS (https://www.iue.tuwien.ac.at/software/viennats/) is described and the adaptations which were made in order to include the ALD process therein are presented. Finally, the implementation of two specific models for ALD of TiO2 and TiN, based on sophisticated surface kinetic models, is provided. 9
PRESENTATION Dr. Véronique Cremers CoCooN Research Group, Ghent University | Post-doctoral researcher Dr. Véronique Cremers already worked at University Ghent from 2014 until 2018. There she studied the conformality of ALD and conformal coating of powders to improve oxidation/corrosion resistance in depth. Afterwards she worked at Plasma Electronic GmbH on the development of ALD/CVD barrier coatings for the packaging industry in Germany. Since 2020 until now she is back in the CoCooN Research Group of University Ghent and deals with the study of conformality of complex ALD coatings. "3D Monte Carlo simulations for a better understanding of conformality of ALD " March 25, 2021 | 16:10 – 16:30 One of the key factors of Atomic layer deposition (ALD) is its capability to deposit highly conformal coatings on 3D (nano)structured surfaces. This feature has rendered ALD a mainstream technique in microelectronics and has triggered growing interest in ALD for a variety of nanotechnology applications, including energy technologies. To get a better understanding of the conformality of ALD and to be able to predict the required exposure for the conformal coating of 3D structures, we developed a 3D Monte Carlo simulation model. Based on this model, we introduced the structure independent concept of Equivalent Aspect Ratio (EAR) which enables a more standardized and direct comparison of reported results concerning the conformality of ALD processes. Other than the conventional aspect ratio, the EAR provides a measure for the ease of coatability by referring to a cylindrical hole as the reference structure. 10
PRESENTATION Dr. Giorgos Gakis National Technical University of Athens | Research Associate Dr. Giorgos Gakis is a chemical engineer, currently working as a research assistant at the National Technical University of Athens (NTUA), Greece. He obtained his PhD from NTUA and Institut National Polytechnique de Toulouse (INPT), France. His main research work is focused on the combined experimental and computational analysis of Atomic Layer Deposition (ALD) and Chemical Vapor Deposition (CVD) processes, for the production of thin films. "Computational modelling of ALD reactors: Linking transport phenomena with surface mechanisms" March 25, 2021 | 16:30 – 16:50 During the last decades, Atomic Layer Deposition (ALD) has emerged as the appropriate process to produce conformal nanometric films, thus answering to the requirements of micro- and opto- electronic devices. The ALD process is complex since it involves transport phenomena occurring inside the reactor chamber and numerous surface phenomena taking place simultaneously in short time scales. This makes the understanding of the ALD process a challenge, due to the difficulty of experimental measurements. In this context, physical based computational modelling has emerged as a tool to provide the required understanding of the different mechanisms and phenomena that take place within the ALD process as well as their effect on the deposited film. The aim of this presentation is to provide a framework for the computational modelling of ALD reactors, combined with experimental analysis of the process, for the case study of alumina ALD. 11
PRESENTATION Jaques Kools Encapsulix | CEO Jaque Kools has 35 years of experience in design, construction and use of vacuum thin film equipment. He has extensive experience in modeling and simulation tools for equipment design, for a number of techniques including ALD, PLD, PVD and IBD. "Simulation for design of ALD equipment" March 25, 2021 | 16:50 – 17:10 In this talk, we will review a number of simulation tools for equipment design: a) Structural, thermal and thermomechanical finite element analysis (FEA) b) Computational Fluid Dynamics (CFD) of the gas flow in the reactor c) Particle-In-Cell (PIC) plasma simulations For each of these techniques, we will discuss features and limitations of the method and illustrate with real world examples. 12
INDUSTRIAL EXHIBITION 13
ORGANIZER European Society of Thin Films Gostritzer Str. 63 01217 Dresden Germany Phone: +49 351 8718370 E-mail: info@efds.org Co-ORGANIZER Fraunhofer Institute for Electronic Nano Systems Chemnitz, Germany BALD Engineering AB Värmdö, Sweden 14
You can also read